Possibly the greatest challenge currently facing the pharmaceutical industry is the development of neuroprotective agents to slow or arrest cell death in neurodegenerative disorders. These disorders involve the death and dysfunction of neurons and of astrocytes. A huge worldwide effort has been directed towards the development of neuroprotective agents that slow or arrest cell death, (Dragunow et al, 2008). Parkinson's disease (PD) is a progressive neurodegenerative disorder that was first described in 1817 by James Parkinson. It is believed to affect 1 in 1000 people over the age of 55 and 1% of people over 65. The clinical features of PD include muscle tremor, muscle rigidity, bradykinesia and impaired balance. There are many theories as to the aetiology of PD but most scientists agree that outside of the rare familial cases, this disorder involves interactions between genetic and environmental factors, (Kennedy et al, 2003). The primary neuropathological feature of PD is the degeneration of the neural connection of the nigrostriatal dopaminergic pathway, between the substantia nigra and the striatum. Dopamine is one of the primary neurotransmitters in the brain and thus the progressive degeneration of the nigrostriatal results in profound dopamine deficiency, with clinical signs of PD appearing when striatal dopamine is reduced by about 80%, (Greenamyre et al, 1993).
Another prominent neuropathological feature of PD is the formation of Lewy bodies. These are abnormal aggregates of the protein alpha-synuclein associated with other proteins such as ubiquitin, neurofilament protein alpha-b-crystalline. They develop inside nerve cells and are believed to be present as an aggresomal response in the cell and thus may play a central role in the disease pathogenesis. Oxidative stress is now also well documented to occur in PD pathogenesis. The inefficient transfer of electrons in the electron transport chain (ETC) of the mitochondria results in the increased production of free radicals, including reactive oxygen species. These free radicals produce oxidative damage by reacting with DNA, lipids and proteins. Oxidative stress-related changes have been detected in the brains of PD patients, (Jenner et al., 1998), and increased oxidative stress has been shown to reduce complex 1 activity of the Mitochondrial ETC.
Most of the drugs currently available to treat PD aim to replace or mimic the actions of dopamine in the brain. The current ‘gold standard' treatment for PD is the drug Levodopa. L-DOPA, the active component of Levodopa, can cross the blood-brain barrier unlike dopamine itself. Once inside the Central Nervous System it is converted into dopamine by the actions of the enzyme DOPA-decarboxylase and thus increases the concentrations of dopamine in the CNS. Levodopa is more effective when it is co-administered with peripheral DOPA decarboxylase inhibitors such as Carbidopa (Sinemet TM) or Benserazide (ProlopaTM). These co-administered drugs stop L-DOPA from being converted to dopamine in the periphery where it is not needed and thus increases the half-life of L-DOPA in the brain from 50 minutes to 1.5 hours. They also help to reduce the side effects caused by dopamine on the periphery. Recently COMT inhibiting drugs have become available to treat PD. These drugs block the actions of the enzyme Catechol-O-Methyltransferase which is responsible for most of the conversion of dopamine outside of the brain. Entacapone is an example of a COMT inhibitor. Dopamine agonists are also now available, such as Ropinirole and pramipoxole. These drugs imitate the actions of dopamine in the brain, without needing to be converted to dopamine itself. Other drugs are available to treat the symptoms of PD, such as anticholinergic drugs or antihistamines.
Ideally the best way of developing therapeutics for human disease is to test them against the disease of interest. However the screening and initial testing of novel compounds directly in humans is generally deemed inappropriate the Federal Drugs Authority and other regulatory authorities. Therefore the use of animal models has become an important tool in the study of the pathogenic mechanisms and therapeutic strategies of human diseases. However animal models are only valuable if they can accurately simulate the pathogenic, histological, biochemical and clinical features of the disease that the investigator wants to examine. Thus an animal model is defined as an animal that has an injury or condition similar to a human disease.
In terms of PD, investigators tend to rely heavily on rodents to model the features of PD and to provide an insight into the mechanisms underlying the pathophysiology of the disease. The first animal model of PD was discovered tragically and accidentally. A group of drug addicts in California in the 1980's consumed a contaminated and illicitly produced batch of the synthetic opiate MPPP (1-methyl-4phenyl-4-propionoxypiperidine). It was contaminated with MPTP (1-methyl-4phenyl-1, 2, 3, 6-tetrahydropyridine). The group developed parkinsonian like symptoms and this brought MPTP forward as a possible cause of Parkinson's symptoms. The book ‘The case of the frozen addicts' by William Langston documented the tragedy. However this tragic incident lead to the development of the first animal model of PD. Over the recent decades many more pharmaceutical agents and environmental toxins have been produced to develop PD models, the most popular of which are 6-hydroxydopamine (6-OHDA), rotenone, paraquat, resperine, methamphetamine and 3-nitrotyrosine models. Most of them have the potential to inhibit complex 1 or to enhance the production of reactive oxygen species through their effect on the mitochondria. The aim of these chemical compounds is to disrupt the nigrostriatal dopaminergic pathway and thereby mimic the striatal dopamine deficiency observed in PD patients. There is controversy as to which models best replicate the progressive nature of PD and an emphasis of recent research has been on the creation of models where exposure is chronic and damage occurs progressively to mimic human PD. There is also doubt as to whether or not an animal model can demonstrate the important distinctions between ‘preclinical' and ‘clinical' disease states. No single model is likely to useful for studying pathogenic mechanisms and for testing therapeutic strategies. However several models are able to replicate one or more of the stages of PD.
Neurotoxin-induced Models Of Parkinson's Disease
1-methyl-4phenyl-1, 2, 3, 6-tetrahydropyridine Model
MPTP is a neurotoxin that causes permanent symptoms of PD by causing nigrostriatal dopaminergic degeneration in the brain. This has lead to it becoming one of the most widely used toxins to mimic the hallmarks of PD. Administration of MPTP results in clinical symptoms remarkably similar to sporadic PD in humans, (Langston et al, 1999). MPTP is not toxic itself but when it crosses the blood-brain barrier it is metabolized glial cell enzyme Monoamine-oxidase B, to yield the toxic cation 1-methyl-4-phenylpyridinium (MPP+). MPP+ is selectively taken up into dopaminergic neurons via its affinity for the dopamine transporter and is thus selectively toxic to dopamine neurons, (Javitich et al, 1985). MPP+ interferes with complex 1 of the electron transport chain, a component of mitochondrial metabolism, leading to cell death and oxidative stress. Therefore MPTP models have been most useful in studies of the molecular changes that underlie mitochondrial dysfunction, (Dawson et al, 2002).
The susceptibility to MPTP varies across species and strains. Rats are not susceptible to the toxin and MPTP potency also varies greatly among mouse strains, (Sedelis et al, 2000). On the whole the MPTP mouse models are disappointing as the DA neurons die rapidly following acute drug administration and thus fail to mimic the progressive nature of PD. More chronic MPTP treatment has resulted in the recovery of the motor behaviour deficits in monkeys once the treatment is stopped. However the MPTP model has played a vital role in studying the mechanism of PD pathogenesis and has also been useful in the testing of PD pathogenesis and of potential neuroprotective therapies.
6-OHDA is a neurotoxin commonly used to lesion dopaminergic pathways and generate experimental models of PD. It was the first chemical agent discovered that had specific neurotoxic effects on the catecholaminergic pathways, (Sachs et al, 1975). Since 6-OHDA is structurally similar to dopamine and noradrenaline and a high affinity for the plasma membrane transporters of these catecholamines, it enters the neurons via these transporters. However the cellular mechanism of 6-OHDA induced neurodegeneration is not well defined. Systemically administered 6-OHDA is unable to cross the blood-brain barrier, but when it is administered directly into the brain it kills DA neurons and terminals, (Javoy et al, 1976). Following stereo-tactical injection of 6-OHDA into the substantia nigra, the dopaminergic neurons start to die in the first 24 hours and striatal dopamine is depleted 2-3 days later, (Hallman et al, 1985). Rats are most commonly used as 6-OHDA models due to their relatively low maintenance costs and established stereo-tactical techniques. Even though the 6-OHDA model has been used to measure the efficacy of antiparkinsonian compounds, (Schwarting et al, 1996), it is still not a completely accurate model of PD as it does not affect other brain regions like human PD and it does not result in the formation of Lewy bodies.
Paraquat and Maneb Model
Paraquat (N, N'-dimethyl-4, 4'-bipyridinium dichloride) is one of the most widely used herbicides in the world, and has emerged as a potential risk factor for PD, based on the fact that it is structurally similar to MPTP's active metabolite MPP+. Paraquat can cross the blood-brain barrier slowly and once inside the cell paraquat is transported into mitochondria, where it is reduced by complex 1 and forms a paraquat radical capable of oxidatively damaging the mitochondria. As of yet however, studies have failed to show progressive DA cell loss or motor deficits, with paraquat treatment.
However when paraquat is combined with maneb, a fungicide that inhibits glutamate transport and disrupts DA uptake and release, (Vaccari et al, 1998), greater effects are seen on the dopamine system than with paraquat treatment alone, (Thiruchelvam et al, 2000). The co-administered paraquat and maneb model have effectively mimicked different stages of PD, including the loss of DA neurons and motor impairments, which adds support to the theory that environmental factors could play a role in PD. However this model has not yet been extensively investigated and it will only ever prove to be an accurate model of PD caused by environmental factors, and not of true PD.
Rotenone is a natural insecticide and pesticide, found in the roots and stems of several plants. It is often used in fisheries to remove unwanted fish species. It is lipophilic and thus can cross the blood-brain barrier and it works by inhibiting the transfer of electrons from complex 1 to ubiquinone in the electron transport chain and thus prevents NADH from being converted into ATP, which can be used as cellular energy. The advantages of the rotenone model are that it results in the specific, chronic and progressive degeneration of the nigrostriatal pathway, following the systemic inhibition of complex 1. The rotenone model also produces formations of ubiquitin and a-synuclein which are similar to the Lewy bodies seen in PD. However there is a high morbidity rate among rotenone treated animals and thus this model is highly labour intensive.
Homykiewicz et al, 1966, found that systemic administration of resperine results in motor symptoms of PD due to striatal dopamine depletion and the loss of dopamine storage capacity in intracellular vesicles. There are major limitations of the resperine model as resperine does not cause morphological changes in the DA neurons of the substantia nigra and the resperine induced changes are only temporary.
As is the case with resperine, METH administration also results in dopamine depletion in the DA nerve terminals due to its neurotoxic effects at high doses in rodents and non-human primates, (seiden et al, 1975). METH is believed to act via the DA receptor and transporter as selective antagonists can block its toxicity, (Schmidt et al, 1985). Unfortunately the METH model is another example of an acute model of DA depletion, which does not reproduce the histological changes seen in human PD, like neuronal DA degeneration and the formation of Lewy bodies.
The 3-NT model was developed to further understand the role of oxidative stress in PD pathogenesis, (Betarbet et al, 2002). It was found that the stereo-tactical injection of 3-NT into the striatum of mice caused a significant reduction tyrosine hydroxylase-positive terminals as well as the loss of substantia nigral DA neurons and motor abilities. Even though it was shown that 3-NT can cause neurodegeneration in animal models, there are still many limitations of the 3-NT model, including its failure to mimic the progressive nature of PD and a lack of protein aggregations and cellular inclusions associated with PD.
Genetic models of Parkinson's Disease
The vast majority of patients with PD are of the sporadic form. However, a person suffering from PD is more likely to have relatives that also have PD. The inheritance of PD is usually quite complex and it is often marked by an earlier onset of PD. A number of specific genetic mutations causing PD have been discovered, which provides an opportunity to generate animals that replicate the pathological and phenotypical features of PD more accurately. As of 2008, genes identified included alpha-synuclein (SNCA), parkin (PRKN), leucine-rich repeat kinase 2 (LRRK2), PINK 1 and DJ1, (Davie et al, 2008).
Alpha-Synuclein Over-Expressing Mice Model
a-synuclein is widely expressed in the nervous system and it plays a role in the modulation of synaptic function in striatal dopaminergic terminals. It is also the primary structural component of Lewy body fibrils. Mutations in a-synuclein have long been associated with familial PD, (Kruger et al, 1998). Two rare mutations have been found in a-synuclein in the genetic form of PD. In the A53T mutation, theonine substitutes for alanine at the residue 53 position. In the A30P mutation phenylalanine substitutes for alanine at the residue 30 position. Feaney and Bender made the first attempts to model PD using a-synuclein. They overexpressed human wild-type and mutant (A53T and A30P) a-synuclein in the fruit fly Drosophila melangaster. They found that DA neurons showed selective depletion after 30-60 days and that motor function was reduced in fly's 45-days and older. The drosophila model is limited however as it lacks glycosylated a-synuclein and only a few locomotor activities can be measured in a fly.
Over the past few decades many lines of mice expressing mutations in a-synuclein have been developed, (Chesselet et al, 2007). These models differ in the promoter used to produce results and this is critical in determining the relevance of the resulting line in modelling the disease. Examples of these promoters include platelet-derived growth factor-b (PDGF-b), thymus cell antigen (Thy 1) regulatory cassette, and the prion-related protein (PrP) promoter. However none of these mice lines show all the features of human PD and only a few have proved useful. For example the inclusions observed when using the PDGF- b growth factor lack fibrillary organisation characteristic of the Lewy bodies observed in PD. Also none of the transgenic mice models exhibit cell death, they only exhibit loss of DA neurons or DA terminals.
Parkin is a protein which in humans is encoded by the PARK 2 gene. Its function is still not full known, but mutations of the gene encoding parkin causes recessively inherited Parkinsonism, (Kitada et al, 1998), usually in PD patients where the age of onset is under 30. The Drosophila fly has also been used to investigate parkin. Yang et al, co-expressed human parkin and a-synuclein in the fruit fly and found that in the presence of parkin, 50% of the DA neurons were protected from death when compared to flies that expressed a-synuclein only. They also had no granular inclusions and had reduced ubiquitin in DA neurons. However the data has yet to show a functional link between parkin and a-synuclein in causing the relatively specific degeneration of DA neurons in PD.
PINK 1 Model
Kowall et al, 2000, has shown that PINK 1 knockout mice show a decrease in evoked DA release in the striatum.
Bernheimer et al, 1973 found that DJ 1 mutations decreased resistance to oxidative stress in cells, flies and mice. However DJ1 mice do not develop DA cell loss.
LRRK 2 Model
Leucine-rich repeat kinase 2 is a protein member of the keucine rich repeat kinase family which in humans is encoded by the LRRK 2 gene. It appears that a late-onset familial PD can be caused by a mutation in the LRRK 2 gene and four LRRK 2 gene variants are found in one third of all PD cases, but infrequently in the general population. Transgenic mouse models are currently being developed.
All of the drugs that are currently available to treat PD are palliative in nature. They work in various ways to reduce the severity of the symptoms of the disease, relieve suffering and improve the quality of life for the people suffering from this serious and complex disease, and some of the drugs are very effective at doing this. However, so far no drugs have been discovered that will halt, delay or reverse the progression of the disease itself or provide a cure for PD. The pharmaceutical industry has spent billions trying to develop a neuroprotective agent that will slow or arrest cell death in neurodegenerative disorders, but as of yet they have failed. Most of the studies appear to fail at the preclinical drug development stage, suggesting a failure of animal models and this was supported in the papers that I reviewed. The first major failures of the animal models is that there is no complete model of PD that will accurately mimic the hallmark features of PD and thus no single model is suitable for all studies. Several models such as the neurotoxin-induced MPTP and Paraquat models fail to show the progressive DA cell loss seen in human PD. Instead the DA cells die rapidly after the toxin is introduced. Other models such as the Methamphetamine model and the 3-Nitrotyrosine model fail to show the production of protein inclusions, similar to the Lewy bodies seen in human PD. The resperine model only induces temporary changes in DA levels in rodents. As far as the genetic models of PD are concerned, the a-synuclein models show inclusions that lack the fibrillary organisation which is characteristic of Lewy bodies in human PD. In terms of parkin, there is no evidence to link parkin to the neurodegeneration of DA neurons. PINK 1, DJ 1 and LRRK 2 have not yet been extensively investigated.
However, even though no model is perfect, the animal models can overlap and demonstrate many of the pathological features of PD and their use has significantly increased our understanding of PD, and lead to possible strategies for developing neuroprotective or neurorestoration agents. However this brings us to the next problem of animal models which is a failure in the translation of results seen in animal models to clinical trials. There have been many drugs that have shown efficacy in preclinical animal models, but have not shown similar effects in humans. This translation from animal models to clinical trials has proven to be very effective in other human disease such as heart disease, cancer and psychiatric illnesses, but it has, as of yet, proved to be largely unsuccessful in neurodegenerative disorders. One possible reason for this failure is that many of the drugs have been tested in young animals or animals with a short lifespan, such as mice and rats. However, most neurodegenerative disorders mainly occur in aged human brains. Also the human brain is a very complex organ that contains at least four distinct types of astrocytes, whereas the rodent models only have two distinct astrocytes, so testing drugs on cultures of rodent astrocytes may lead to inaccurate results. For this reason, it is thought that non-human primates are likely to be the most relevant models of PD, due to their greater brain capacity, longer life-span and similar pharmacokinetic properties to humans. Other possible reasons for the failure of translation from animal models include: an inability of compounds to cross the blood-brain barrier, inaccurate engagement of the drug target in humans, toxicity of compounds and pharmacokinetic and pharmacodynamic problems.
Due to the increasing age of the population as a whole and thus the increasing incidences of neurodegenerative disease, there is more of an incentive for biotechnology and pharmaceutical companies to develop, effective neuroprotective agents to be used in humans. However if they are to be successful they must overcome the failures of the current animal models. They must develop a model that can accurately mimic all of the hallmark features of human PD, or improve on the translation of the current animal models from the preclinical stage to the clinical stage of drug development, possibly through improving target validation.